Meeting global energy demand using intermittent renewable sources, particularly wind, remains a significant challenge due to weather variability. Reliable wind energy forecasting requires accurate weather predictions and detailed microscale modeling of wind-turbine interactions. To address this, a coupled framework combining the Weather Research and Forecasting model (WRF) [1] and the Computational Fluid Dynamics (CFD) solver, Xcompact3D [2], with data exchange facilitated by the Multiscale Universal Interface (MUI) [3, 4], is being developed. WRF, a versatile numerical weather prediction system, provides initial and boundary conditions for Xcompact3D, which is a high-order CFD solver effectively applied to wind-farm simulations and atmospheric studies. The header-only MUI library enables efficient data exchange between solvers via the Message Passing Interface. This open-source framework is optimized for scalability on high-performance computing (HPC) systems. The framework allows one- and two-way coupling between multiple instances of Xcompact3D and WRF. This capability can support real-time simulations with realistic boundary conditions, enabling the creation of digital twins for wind farms or urban areas. The framework also facilitates local grid refinement in Xcompact3D, enhancing modeling accuracy. The framework was validated using the Askervein Hill experimental campaign [5], which recorded air velocity and direction data. Validation involved six nested domains: WRF used four domains with the coarsest resolution of 13.5 km, and Xcompact3D used two domains with the finest resolution of 11.5 meters. Results demonstrate accurate simulation of velocity fields and scalability of the computational framework on HPC systems, shown in figures 1 and 2.